1. Field of the Invention
This invention relates to compositions and methods for the treatment of viral disorders. Treatment involves administration of an inducing agent, to induce expression of a product in a virus-infected cell, and an anti-viral agent, that acts on the expressed product to destroy the virus-infected cell.
2. Description of the Background
A growing number of cellular disorders such as neoplastic malignancies have been found to contain viral genetic sequences or virus particles in the anomalous cells. For a large number of these disorders, the presence of the virus is believed to be causative or at least contributory instrument. Representative members of many of the known families of viruses have been found in such cells including members of the herpes family of viruses, the polyomaviruses and the hepatitis viruses.
Epstein-Barr virus (EBV), a 172 kb herpes virus, is often found intimately associated with both mature and immature B cells and is believed to be involved to some degree in infectious mononucleosis, African Burkitt's lymphoma (BL) and nasopharyngeal carcinoma. EBV undergoes lytic replication after initial infection of oropharyngeal epithelia. The linear form genome is duplicated, packaged into the viral capsid and extruded from the cell by budding or lysis. One hundred viral proteins are synthesized during this lytic stage of the virus life cycle. In contrast, normal B cells incubated with EBV in vitro are efficiently immortalized and develop into continuously growing lymphoblastoid cell lines (LCLs). The cellular events that regulate these distinct outcomes are as yet unclear.
In immortalized cells, the genome circularizes, amplifies and replicates coordinate with, and dependent upon, cell division. Because no viral particles are produced, infection is considered to be latent and EBV persists in the cells for life.
Outgrowth of latently infected B cells is prevented by T cell immune surveillance. In immortalizing latent infection, only 11 gene products are detected, including 6 nuclear antigens (EBNA-1, -2, -LP, -3A, -3B, -3Q), 3 membrane proteins (LMP-1, LMP-2A, LMP-2B) and two small, non-poly(A) RNAs (EBER-1 and EBER-2) (G. Miller et al., Epstein-Barr Virus: Biology, Pathogenesis and Medical Aspects, Raven Press, N.Y. 1990). In EBV(+) tumors such as Burkitt's lymphoma, neoplastic genetic events have often superseded the requirement for viral immortalizing functions, and gene expression may be limited to EBNA-1 (M. Rowe et al., EMBO J. 6:2743-51, 1987). Virus tropism is determined by complement receptor type 2 which mediates attachment of the envelope protein gp350/2204 to B and some T lymphocytes, follicular dendritic cells and epithelial cells.
EBV is a common and worldwide pathogen. Childhood infection is asymptomatic. About 50% of individuals with delayed exposure develop a self-limited, lymphoproliferative syndrome referred to as infectious mononucleosis. EBV is also detected in 2 endemic tumors, African Burkitt's lymphoma (BL) (G. Henle et al., Proc. Natl. Acad. Sci. USA 58:94-101, 1985) and nasopharyngeal carcinoma (NPC) (W. Henle et al., Adv. Viral Onco. 5:201-38, 1985), as well as gastric carcinoma, breast cancers and sarcomas. Recently, some T-cell and B-cell lymphomas, as well as about 50% of Hodgkin's lymphomas have been found to contain EBV (L. M. Weiss et al., N. Engl. J. Med. 320:502, 1989).
Burkitt's Lymphoma occurs in the United States at a prevalence of about 60 plus EBV patients per year and can occur several years after the primary EBV infection in immunocompetent patients. Burkitt's lymphoma is a monoclonal lymphoma, as opposed to infectious mononucleosis, which is a polyclonal disease. The disease is distinctly different in Africa, where it is endemic where about 97% of the population is EBV-positive with 100 sero-positives per 10.sup.6 children. In the U.S., infection is sporadic with about 20% EBV-positives and somewhat rare with 1-2 per 10.sup.6 sero-positive children or about 300 cases per year.
African Burkitt's lymphoma is characterized by rapid growth of the tumor at non-lymphoid sites such as the jaw or the retroperitoneum. The tumor is of B cell origin and is closely related to the small noncleaved cells of normal lymphoid follicles. Biopsy specimens from African Burkitt's lymphoma invariably contain the EBV genome and are positive for EBNA (I. Magrath, Epstein-Barr Virus and Associated Diseases, pp. 631-43, M. Ninjhoff Publishing, Boston, 1986). This contrasts with the non-African Burkitt's lymphoma, in which only 15% to 20% of the tumors contain the EBV genome. EBV has a worldwide distribution and infects most (more than 90%) individuals before adulthood. The clustering of Burkitt's lymphoma in the equatorial belt of East Africa remains unexplained. It has been hypothesized that alterations of the immune system, possibly due to hyperstimulation by endemic malaria, may play an important role in the outcome of an EBV infection to individuals in this region (D. J. Moss et al., Int. J. Cancer 31:727-32, 1983). Individuals from this region show impairment in virus-specific cytotoxic T-cell activity. Normally, it is the T-cell response to EBV infection that limits B-cell proliferation, and this T-cell response is directly stimulated by EBV (H. zur Hausen et al., Nature 228:1056-58, 1970). It has been postulated that the failure of the T-cell immune response to control this proliferation could lead to excessive B-cell proliferation and, as such, provide a suitable background for further mutation, oncogenic transformation, and lymphomagenesis.
A scenario has been suggested for the involvement of EBV in the etiology of African Burkitt's lymphoma (G. Klein, Proc. Natl. Acad. Sci. USA 76:2442-46, 1979). The first step involves the EBV-induced immortalization of B lymphocytes in a primary infection. The second step involves the stimulated proliferation of EBV(+) B cells. This step is facilitated in the geographic areas where Burkitt's lymphoma is endemic (presumably because of the presence of malaria), through B-cell triggering and the suppression of T-cells involved in the control of the proliferation of EBV-infected cells. This pool of cells becomes increased in size as a target cell population for random chromosomal rearrangements. The third and final step is the reciprocal translocation involving a chromosomal locus with an immunoglobulin gene and the c-myc gene on chromosome 8. This leads to the deregulation of the c-myc gene, to the development of the malignant clone and to the appearance of a tumor mass (G. Klein et al., Nature 315:190, 1985). Alternative scenarios have been proposed in which the order of the steps are rearranged such that the B-cell activation by malaria precedes the chromosomal translocation and is followed by EBV infection. Regardless, the components of these two scenarios each account for the geographic distribution of Burkitt's lymphoma, the critical involvement of EBV in lymphomagenesis, and the eventual selection and clonal outgrowth of a population of cells with the critical translocation involving the deregulation of the c-myc gene on chromosome 8.
Treatment of Burkitt's lymphoma is most commonly chemotherapy, with radiotherapy playing a minor role, With this regiment, prolonged survival rates of 50% can be achieved.
Hodgkin's disease has an incidence in United States of about 3,500 EBV(+) patients per year. For more than 20 years, a role for EBV in the pathogenesis of Hodgkin's disease (HD) was postulated based on epidemiologic evidence linking Hodgkin's patients with EBV seropositivity and elevated EBV titers (A. S. Evans et al., Int. J. Cancer 34:149, 1984). A number of studies have found an increase (2-5 fold) in the incidence of HD after infectious mononucleosis. However, some Hodgkin's patients were seronegative for EBV and the association between EBV and Hodgkin's disease remained speculative until 1987. In that year, molecular genetic analysis demonstrated that some Hodgkin's tissues contained monoclonal EBV DNA and that the virus was localized to Reed-Sternberg (RS) cells (the malignant cells in HD). Subsequent immunohistochemical and serologic data support an association between EBV and Hodgkin's disease and confirm the localization of the virus to cytologically malignant-appearing RS cells and variants (N. M. Jiwa et al., Histopathology 21:51, 1992). EBV also infects variable numbers of small B and T lymphocytes in the reactive inflammatory cell infiltrate that composes the bulk of Hodgkin's tissues (L. M. Weiss et al., Am. J. Path. 139:1259, 1991).
In most series of cases reported to date, EBV is associated with approximately half of mixed cellularity Hodgkin's disease cases and a somewhat lower percentage of the nodular sclerosing subtype. In contrast, lymphocyte-predominant Hodgkin's disease is rarely EBV-associated. Although EBV DNA has been identified in the lymphocyte depletion variant of Hodgkin's disease, there is controversy over the morphologic criteria for diagnosis of this variant and its distinction from anaplastic large-cell lymphoma, in which EBV is also implicated as a pathogen.
Clonal and non-clonal EBV genomes are present in Hodgkin's disease. Expression of the oncogene LMP (latent membrane protein) is seen in RS cells. In HD, the region of the (viral) BNLFI oncogene coding for the amino terminal and transmembrane domains (associated with oncogenic function) of LW appears to be homogeneous, whereas the region coding for the intracytoplasmic (carboxyl terminal) domain of LMP is heterogeneous. Cytological similarities between RS cells and immunoblasts of known EBV-induced infectious mononucleosis and EBV induced AIDS-related lymphomas are consistent with the hypothesis that the EBV-BNLFI, oncogene is an inducer of morphological features of RS cells. Whether chromosomal integration of EBV DNA is an important factor in activation of such a transforming activity remains to be elucidated. Therefore the RS cells appear to be derived from lymphocytes beyond the pre-B-cell or common thymocyte stage, which may or may not subsequently become infected by EBV.
The high prevalence of EBV in Hodgkin's disease implies an etiologic role for the virus in Hodgkin's tumorigenesis. This pathogenetic theory is supported by the monoclonality of EBV DNA in these tumors (M. L. Gulley et al., Blood 83:1595-602, 1994). In one series, monoclonal EBV DNA was detected in all 17 cases having EBNA1-positive RS cells. Because tumor-associated viral DNA is monoclonal, it is likely that virus infection preceded clonal expansion. This reinforces the hypothesis that the virus is not an innocent bystander, but rather plays a role in the pathogenesis of the Hodgkin's disease and the other tumor types in which it is found (A. Neri et al., Blood 77:1092, 1991). The observation of EBNA1 expression in the RS cells of clonally-infected cases indicates that the clonal virus is localized to these cells and suggests that Hodgkin's disease results from the transformation of an EBV-competent cell. Other studies suggest that this virus is modulating rather than an etiologic agent in a considerable proportion of HD cases.
Investigations into the biology of EBV infection have shown that only one viral particle successfully infects a given cell. Once the viral DNA is established inside the cell, it circularizes and reproduces itself to yield multiple identical copies of viral DNA (E. A. Hurley et al., J. Exp. Med. 168:2059, 1988). In this way, tumors derived from infected cells can have multiple copies of EBV per cell, while maintaining clonal viral DNA structure. The average amount of clonal EBV DNA in Hodgkin's disease tissues varied from 0.5 to 5 copies per cell. Because RS cells comprised only a small fraction (&lt;1%) of all HD tissue cells, the content of EBV DNA in each RS cell is estimated to be at least 100 times higher than the measured average copy number per cell, or at least 50 copies of viral DNA per RS cell. This is comparable with, or greater than, the viral burden in infected non-Hodgkin's lymphomas. The high copy number of EBV in RS cells may relate to the pathobiology of this complex lymphomatous disorder. In agreement with these studies, EBV DNA is abundant and monoclonal in infected RS cells. The presence of EBV in RS cells was strongly and independently linked to mixed cellularity histology and Hispanic ethnicity.
T-Cell non-Hodgkin's lymphoma (NHL) has a U.S. incidence of about 3,100 T-cell, EBV(+) NHL patients per year. Clonally-integrated EBV is found in association with T-cell lymphomas as well as B-cell lymphomas. Currently, three populations of tissue-restricted T lymphocytes have been recognized, mucosa-associated, cutaneous and nodal T lymphocytes. T-cell lymphomas arising from different sites, but with similar morphology may show differences in lymphomagenesis and in expression of oncogenies, adhesion molecules, presence of certain DNA/RNA viral sequences and in clinical presentation and behavior.
Primary cutaneous CD30(+) large cell, T-cell lymphomas often remain localized to the skin for a long time, express a unique cutaneous lymphocyte antigen (CLA), known as the skin-homing receptor, have been postulated to be associated with the presence of human T-cell leukemia/lymphoma virus type I (HTLV 1), and have a good clinical course (R. C. Beljaards et al., Cancer 71:2097, 1993). In contrast, morphologically similar T-cell lymphomas of nodal origin often behave more aggressively, are CLA-negative, and have been associated with the presence of EBV (P. C. de Bruin et al., Histopathology 23:127, 1993). There was no relation between primary cutaneous T-cell lymphoma and EBV.
In agreement with this finding, EBV is not present in lymphomatoid papulosis, a premalignant cutaneous lymphoproliferative disorder (M. E. Kadin et al., J. Pathol. 170:145, 1994). EBV-associated T-cell lymphomas are highly site-restricted and are morphologically indistinguishable from EBV(-) T-cell lymphomas (P. C. de Bruin et al., Blood 83:1612, 1994).
Nasal T-cell lymphomas are EBV-associated (F. C. S. Ho et al., Hematol. Oncol. 8:271, 1990). The frequency of EBV(+) primary pulmonary T-cell lymphoma is similar to the frequency of EBV(+) primary nodal T-cell lymphoma (J. Sabourin et al., Am. J. Surg. path. 17:995, 1993; P. C. de Bruin et al., Histopathology 23:509, 1993). EBV-associated primary gastrointestinal T-cell lymphomas seem to be rare, but recently a considerable number (36%) of enteropathy-associated T cell lymphoma cases (EAT CL) were reported to be EBV(+) (L. Pan et al., Pathology 170:137, 1993), although other studies did not find such an association. Angiocentric immunoproliferative lesions (AILs) of nose (also known as lethal midline granuloma), lung (also known as lymphomatoid granulomatosis), and skin are frequently associated with EBV (L. J. Medeiros et al., Am. J. Surg. Pathol. 16:439, 1992). Although all AILs in the nose were EBV-positive, the relation between angiocentricity and EBV seems to be more complex. Primary cutaneous and gastrointestinal AILs are EBV(-), whereas primary pulmonary AILs can be EBV(+) or EBV(-). There is no clear relation between angiocentricity and EBV, and only primary site seems to be important in the relation between peripheral T-cell lymphoma and EBV. Thus, there are site-restricted differences in the occurrence of EBV-infected peripheral T-cell lymphomas.
Infection of T cells by EBV most likely occurs via CR2 or CR2-like receptors (C. D. Tsoukas et al., Immunol. Today 14:56, 1993). The close contact between T cells and the upper respiratory tract epithelium, known for its reservoir function for EBV, probably make T cells in this region more vulnerable for EBV infection. The finding that EBV can be found in almost all tumor cells in most cases of primary extra-nodal, and especially nasal, T-cell lymphoma, in contrast to primary nodal T-cell lymphoma, where the number of EBV-infected neoplastic cells varies greatly between the cases argues for an etiologic role for EBV in these cases. Moreover, these cases often express LMP-1, known for its transforming and oncogenic properties in vitro and are reported to be monoclonal for EBV (I. Su et al., Blood 77:799, 1991). Thus, there are site-restricted etiologic differences between morphologically identical T-cell lymphomas, of which EBV might be one of many factors.
EBV-induced lymphoproliferative disease or lymphoma has an immunodeficiency incidence in U.S. of about 10,000 B-cell, EBV(+) lymphoma patients per year. EBV is very commonly associated with lymphomas in patients with acquired or congenital immunodeficiencies. These lymphomas can be distinguished from the classical Burkitt's lymphomas in that the tumors may be polyclonal. Tumors also do not demonstrate the characteristic chromosomal abnormalities of Burkitt's lymphoma described earlier. The pathogenesis of these lymphomas involves a deficiency in the effector mechanisms needed to control EBV-transformed cells. The rototypic model for this disease has been the X-linked lymphoproliferative (XLP) syndrome (D. T. Purtilo et al., Am. J. Med. 73:49-56, 1982). Patients with XLP who develop acute infectious mononucleosis exhibit the usual atypical lymphocytosis and polyclonal elevation of serum immunoglobulins and increases in specific antibody to VCA and to EA. During these infections, patients with XLP fail to mount and sustain an anti-EBNA response after acute EBV infection. The unique vulnerability of males with XLP to EBV infection is most likely due to an inherited immune regulatory defect that results in the failure to govern the cytotoxic T cells and NK cells required to cope with EBV.
Patients with iatrogenic immunodeficiencies, such as organ transplant recipients, are also at an increased risk for lymphomas, and these lymphomas often contain EBV DNA and EBNA. Also, patients with AIDS are at a higher risk for developing polyclonal lymphomas associated with EBV.
EBV-induced lymphomas are associated with immunosuppression. EBV-associated lymphoproliferative disease (EBV-LPD) is characterized by actively proliferating EBV(+) B-cells, frequently without overt malignant change. These immunoblastic lymphoma-like lesions have been identified in a variety of transplant patients, in patients with congenital immunodeficiency, and in patients infected with HIV (J. I. Cohen, Medicine 70:137-60, 1991). These so called post-transplant lymphoproliferative disorders (PTLD) are observed after all transplants, including kidney, bone marrow, heart, liver and lung transplantation. This increased incidence of EBV-LPD in this setting is likely due to the aggressive immunosuppression required after these transplants.
Phenotypically, immunoblastomas resemble large cell lymphomas (LCLs) in vitro. They express EBNA1-EBNA6 and LMP. Obviously, these cells proliferate because the surveillance of the host has broken down, rather than owing to any cellular escape (G. Klein, Cell 77:791-93, 1994). These tumors share several common features, including rapid progression, clinical aggressiveness and a tendency to grow at extra-nodal sites (E. B. Papadopoulos et al.. N. Engl. J. Med. 330:1185-91, 1994). These EBV(+) B lymphomas generally have an ominous prognosis, especially if monoclonal. In some cases, rapid immune reconstitution has led to spontaneous regression of the tumor. In most cases, however, this has not been possible or is ineffective. EBV-LPD has been particularly tragic in the transplant setting, causing the death of recently "cured" patients who would be unlikely to develop such disease as their natural immune reconstitution occurred over time. Empirically, nonclonal PTLD may regress with decreased immunosuppression (if that is clinically feasible), whereas regression is less predictable in polymorphic, monoclonal tumors. In contrast, monoclonal monomorphic tumors are frequently more aggressive and tend to progress despite immune modulation. Treatment with nucleoside antivirals alone, surgery, cytotoxic chemotherapy, .alpha.-interferon with immune globulin, or B-cell directed complement fixing monoclonal antibodies, has rarely been successful. Recently, autologous, T-cell transfusions were applied with some success in LPD after allo-BMT, but are unlikely to become a routine therapy and would be inappropriate in tumors of host origin.
EBV's association with PTLD is based on several clinical observations. Almost all patients with PTLD have serologic evidence of an EBV infection and the disease involves a proliferative disorder of B cells. The B cell lymphoproliferation can be either monoclonal or polyclonal. EBV-LPD is characterized by latent viral replication. The genome persists as a circular episome within the infected B-cells, and lytic replication is restricted (C. M. Rooney et al., XVIII Intl. Herpesvirus Workshop, abstract B3-37, Jul. 25-30, 1993). Additional evidence that EBV is associated with post-transplant lymphoproliferative disorders is the finding that patients without any prior exposure to EBV are at highest risk for developing the lymphoproliferative disorder. Patients who are serologically EBV(-) are at the highest risk when receiving an EBV(+) transplant. Supporting this observation is the finding that EBV-associated PTLD is more frequent in children receiving transplants than in adults. The incidence of EBV-associated PTLD is three-fold greater in some series of pediatric transplantation. An intriguing report suggested that the EBV is carried with the graft. In this case, a common donor gave rise to PTLD in two separate organ recipients. Both recipients demonstrated the same chromosomal translocation in their tumors.
Lung transplant recipients are a special sub-group of patients at risk for developing PTLD. Most lung transplant recipients require relatively high immunosuppression for prevention of acute and chronic graft rejection. The development of PTLD and these patients has been estimated at 5% to 10%. Clinical presentation has ranged from systemic adenopathy to masses and infiltrates in the transplanted lung. The existing literature again indicates that there two distinct clinical syndromes in lung transplant recipients with PTLD. One scenario involves the polyclonal lymphoproliferative disorder commonly associated with mononucleosis and other benign lymphoproliferative diseases. This syndrome appears to respond to a decrease in the immunosuppressive regimen. The second clinical presentation involves a more aggressive form of lymphoproliferative disorder, one which histologically resembles immunoblastic non-Hodgkins lymphoma (NHL). This EBV-associated lymphoma is often monoclonal and can display an aggressive clinical course. For example, tumors are typically unresponsive to conventional chemotherapy and have been uniformly fatal. Although withdrawal of immunosuppression has been attempted in many cases, the recipients' dependency on graft function for adequate oxygenation complicates any reduction in the immunosuppression regimen.
The development of a large patient population with T-cell dysfunction in the 1980's, caused by profound iatrogenic immunosuppression required to facilitate solid organ and bone marrow transplantation. The AIDS epidemic led to the discovery of 2 additional EBV-related illnesses hairy leukoplakia (J. S. Greenspan et al., N. Engl. J. Med. 313:1564, 1985), and B-lymphoproliferative disease (EBV-LPD) (D. W. Hanto et al., Ann. Surg. 198:365-69, 1983).
As a result of effective antiretroviral therapy, prophylactic measures, and rigorous treatment of opportunistic infections (AIDS-related, EBV-associated lymphomas), HIV-infected individuals now live longer. More than 40% of people with AIDS will develop a neoplastic disease, resulting in severe morbidity and, often, death. Intermediate or high-grade B-cell lymphoma was added as an AIDS-related malignancy in 1985. Prolonged survival in HIV-infected patients is associated with increased incidence of lymphoma. Between 1940 and 1980, the VS incidence of lymphoma. doubled. The AIDS epidemic, however, has superimposed on this underlying trend an additional risk of lymphoma between 60 and 100 times that expected in the HIV-negative population. Lymphoma, is likely to be a late manifestation of HIV infection, as documented in France where 33% of lymphomas occurred after an AIDS-defining illness. In the US, only about 3% of cases of AIDS are diagnosed simultaneously with lymphoma. Lymphoma occurs among all population groups at risk for HIV in all age groups and in diverse geographic regions. In the HIV-negative population, women have a lower incidence of lymphoma than men, and the incidence increases with age in homosexual/bisexual men and in hemophiliacs. The same incidence patterns are seen in the HIV-positive population, except among intravenous drug users, who may succumb relatively early to various infections.
Eighty to ninety percent of AIDS patients who are diagnosed with lymphomas are either intermediate or high-grade B-cell tumors, such as immunoblastic or large-cell types, or small noncleaved lymphoma, which may be subclassified as either Burkitt's or Burkitt-like. Among HIV-negative lymphoma patients, high-grade lymphomas represent only 10% to 15% of all lymphomas.
Etiology and pathogenesis of lymphomas in patients infected with HIV is multifactorial. Immunosuppression is thought to be a primary factor correlated with an increased incidence of lymphoma in several settings, including certain congenital immune deficiency diseases, autoimmune disorders, and the chronic use of immuno suppressive drugs, as occurs in patients who have undergone organ transplantation. Lymphomas which develop in these settings are similar to the AIDS lymphomas in terms of the pathologic type, the high frequency of extra-nodal disease at presentation, and the relatively poor prognosis. It is clear that EBV plays a major role in AIDS lymphomagenesis. This may perhaps be the result of damaged immunosurveillance of EBV-infected cells in AIDS patients.
There appears to be some biological significance to the site of disease. Patients with lymphoma primary to the central nervous system have an extremely poor prognosis and progressive HIV infection. About 75% have a history of AIDS as well as low CD4 cell counts (&lt;50 mm.sup.3) prior to development of lymphoma. Pathologically, these primary central nervous system lymphomas are almost always immunoblastic or large-cell lymphomas and uniformly associated with EBV. In contrast, improved prognosis has been associated with the large-cell type of systemic lymphoma, exclusive of the central nervous system. Clinicopathologic correlations are imprecise at best, because patients who develop lymphoma as the first AIDS-defining diagnosis may be distinct from those who develop lymphoma after preceding opportunistic illness.